184
Biochemistry 2003, 42, 184-190
Remarkable Stabilization of Neutrophil NADPH Oxidase Using RacQ61L and a p67phox-p47phox Fusion Protein† Kei Miyano, Hiroyasu Fukuda, Kentaro Ebisu,‡ and Minoru Tamura* Department of Applied Chemistry, Faculty of Engineering, Ehime UniVersity, Matsuyama, Ehime 790-8577, Japan ReceiVed September 26, 2002; ReVised Manuscript ReceiVed NoVember 12, 2002
ABSTRACT: Activation of the phagocyte NADPH oxidase occurs via assembly of cytosolic p47phox, p67phox, and Rac with the membrane-bound flavocytochrome b558. Recently, we have found that p67phox-(1-210) (p67N) fused with p47phox-(1-286) (p47N) or with Rac efficiently stabilizes the oxidase in a cell-free reconstitution system. In an attempt to further stabilize the oxidase, we herein used a constitutively active Rac, RacQ61L, and examined its effect on the oxidase stability. The half-life (t1/2) of the activity reconstituted with wild-type Rac was 12 min at 37 °C, which was extended 6-fold by RacQ61L. Also, the stability of the oxidase without p47phox increased 8-fold using RacQ61L. RacQ61L had a higher affinity for the complex than wild-type Rac and increased the affinity of p67N for the complex. Far-western blotting showed an enhanced binding between RacQ61L and p67N. The oxidase was stabilized by nanomolar FAD, and RacQ61L lowered the FAD concentration required. The combination of RacQ61L and a fusion protein consisting of p67N and p47N produced an extremely stable enzyme (t1/2 ) 184 min at 37 °C). The effectiveness of RacQ61L and fusion proteins on stabilization was in the following order: p67N-Rac < p67N + RacQ61L e p67N-RacQ61L , p67N-p47N + RacQ61L. These results indicate that a tightly bound ternary complex of p67phox, Rac, and p47phox is very effective in maintaining the oxidase and confirm that the longevity of the activated state requires continuous association of these components. This simple and efficient method of stabilization may provide a useful tool to elucidate the nature of the activated oxidase.
The phagocyte NADPH oxidase is a multicomponent enzyme that produces superoxide anion (O2-) in response to exposure of phagocytes to pathogens such as bacteria or fungi (1). The enzyme is dormant in resting cells and becomes active upon cell stimulation by pathogens or a receptor-mediated stimulant (2, 3). The activation occurs via assembly of the cytosolic regulatory proteins p47phox, p67phox, and Rac with the membrane-associated flavocytochrome b558 (cyt b558)1 (3). Cyt b558 consists of p22phox and gp91phox, the latter of which is classified as Nox-2 in the Nox/Duox family (4). Two other factors, p40phox and rap1A, are also assumed to be involved in the enzyme regulation although they are not essential for the activity. Interactions among the subunit proteins have been extensively studied (5, 6). SH3 (Src homology 3) mediated interactions among p47phox, p67phox, and p22phox have been demonstrated (7-9). An interaction between p67phox and Rac has also been revealed (10, 11), and the interaction was shown to occur between the TPR (tetratricopeptide repeat) domain of p67phox (12, 13) and the switch I region of Rac † Supported by a grant (13480297) for Scientific Research from the Ministry of Education, Science, Sports, and Culture of Japan. * To whom correspondence should be addressed. Tel: 81-89-9279938. Fax: 81-89-927-8546. E-mail:
[email protected]. ‡ Present address: Chemo-Sero-Therapeutic Research Institute, Ohkubo, Kumamoto 860-8568, Japan. 1 Abbreviations: cyt b558, cytochrome b558; RacQ61L, Rac mutant in which Gln-61 is replaced by Leu; p67N, C-terminal-truncated p67phox (residues 1-210); p67N-RacQ61L, fusion protein between p67N and RacQ61L; p47N, C-terminal-truncated p47phox (residues 1-286); p67N-p47N, a fusion protein between p67N and p47N.
(13, 14). However, actual interactions in the active complex and the whole structure of the complex are still sketchy. With regard to the role of cytosolic phox proteins in the activation, the following points have been noted: (i) Rac and p67phox are the minimum activating components (15, 16), (ii) p47phox functions as an adapter protein (15) and a stabilizer (17, 18), (iii) p67phox is involved in the regulation of electron flow from NADPH to FAD (19, 20), and (iv) an activation domain of p67phox is involved in the regulation (21), probably by direct interaction with cyt b558 (20). Despite these findings, the mechanism for activation of the oxidase has remained unclear. The activated NADPH oxidase is highly labile, complicating investigations of the subunit structure and preventing isolation of the active enzyme complex (22). In earlier studies we found that a chemical cross-linker remarkably improves the stability of the oxidase in crude cell-free systems (17). We also applied the methodology to a pure reconstitution system consisting of the recombinant cytosolic components and purified cyt b558, but the effect was not as dramatic as in the crude reconstitution systems.2 Recently, we constructed fusion proteins between Cterminal truncated p67phox (p67N) and p47phox (p47N) and found that p67N-p47N, but not p47N-p67N, remarkably improves the stability of the oxidase activity (23). Subsequently, fusion between p67N and Rac was also investigated, and it was shown that p67N-Rac, but not Rac-p67N, 2
M. Tamura, unpublished results.
10.1021/bi0269052 CCC: $25.00 © 2003 American Chemical Society Published on Web 12/14/2002
Tightly Bound Phox Triad Keeps NADPH Oxidase on considerably improves the stability (18). From these results we proposed a model for the topology of these phox proteins in the active complex (18). Although these fusion proteins reconstitute the oxidase activity that is fairly stable at 25 °C, the activity thus obtained was still not stable enough at higher temperatures (e.g., 37 °C).3 In an attempt to further stabilize the oxidase activity and to investigate the nature of the complex, we have used a constitutively active Rac, RacQ61L, and examined its effect on the stability of the oxidase activity in a cell-free reconstitution system. We find that the RacQ61L mutant remarkably stabilizes the oxidase activity and the effect could be ascribed to an enhanced binding of RacQ61L to p67phox. We applied a fusion technique to the RacQ61L-containing system and found that when used with a fusion protein, p67N-p47N, RacQ61L produces an extremely stabilized oxidase activity. The mechanism for the excellent stabilization is discussed in relation to the topology model of the complex. EXPERIMENTAL PROCEDURES Materials. pGEX-2T, pGEX-6P, Escherichia coli BL-21, and glutathione-Sepharose were purchased from Amersham Pharmacia Biotech (Little Chalfont, U.K.). Oligonucleotide primers were synthesized by the same manufacturer. EcoRI and BamHI were purchased from Toyobo Inc. (Tokyo, Japan). GTP, GTPγS, GDP, and FAD were purchased from Nacalai Tesque (Kyoto, Japan). NADPH was obtained from Oriental Yeast Co. (Tokyo, Japan). Preparation of Rac, p47phox, and p67phox (1-210). Complementary DNAs encoding human Rac [Rac1(C189S)], p47phox, and p67phox-(1-210) (p67N) were gifts from Dr. Dave Lambeth (Department of Pathology, Emory University School of Medicine). The protein Rac1(C189S) is referred to as wild-type Rac in this paper. We constructed the RacQ61L mutant on the basis of Rac 1(C189S), which is fully active and more stable than natural Rac1. The pGEX2T plasmid containing the cDNA for Rac or p67N or the pGEX-6P plasmid containing p47phoxcDNA was transfected into BL-21 cells. The proteins were expressed and purified as previously described (23). Construction of Recombinant Plasmids for RacQ61L and p67N-RacQ61L Fusion Proteins. The Rac cDNA in pGEX2T was mutated to convert glutamine-61 (CAA) to leucine (CTA) using a site-directed mutagenesis kit (Stratagene Quik Change). For the fusion protein p67N-RacQ61L, the pGEX2T containing the p67N-Rac cDNA (18) was engineered to convert glutamine-61 (CAA) to leucine (CTA) in the Rac moiety using the Quik Change kit. The sequences of the mutant and fusion proteins were confirmed by dideoxynucleotide-based sequencing. Expression of Mutant and Fusion Proteins. The pGEX2T containing RacQ61L, p67N-RacQ61L, or p67N-p47N cDNA was transfected into E. coli BL-21, expressed at 37 °C, and lysed using the conditions previously described for Rac (23). In the case of p67N-RacQ61L, the cells were frozen and thawed before lysis to improve the protein recovery as described for preparation of p67N-Rac (18). The expressed proteins were purified with glutathione3
K. Ebisu and M. Tamura, unpublished results.
Biochemistry, Vol. 42, No. 1, 2003 185 Sepharose beads, concentrated with a Centricon Y-10 (Millipore Corp., Bedford, MA), and stored at -80 °C until use. Preparation of Cyt b558. Isolation of plasma membranes from porcine neutrophils and purification of cyt b558 were performed as previously described (18). The cyt b558 preparation obtained was relipidated with a mixture of phospholipids and cholesterol (PC:PE:PI:SM:cholesterol ) 31:14:7.5:23.2: 24.3 wt % of total lipid) and stored at -80 °C until use. Reconstitution of NADPH Oxidase and Assay for O2Generation. Rac (6 µM) was preincubated with 100 µM GTP at 25 °C for 20 min in buffer A (20 mM potassium phosphate buffer, pH 7.0) and reconstituted with p67N (6 µM) with 1.3 µM p47phox and cyt b558 (0.1 µM) in 50 µL of buffer A containing 4 mM MgCl2, 10 µM FAD, and 10 µM GTP. In some experiments, RacQ61L was used instead of Rac without preloading (see Results). The mixture was incubated with 200 µM SDS for 5 min at 25 °C to activate NADPH oxidase. Four 10 µL aliquots of the reaction mixture were then transferred into wells of a 96-well plate and diluted 25-fold with buffer A containing 4 mM MgCl2, 200 µM NADPH, 10 µM FAD, and 80 µM cytochrome c. Superoxide generation was measured by monitoring the cytochrome c reduction at 550 nm using a microplate reader (Tecan, Spectra Classic). The assays were sometimes repeated in the presence of superoxide dismutase (80 µg/mL) to verify O2--dependent cytochrome c reduction, and the rate of O2- generation was corrected. GTPase ActiVity. GTPase activities of Rac and its mutants were measured by a filter binding assay using [γ-32P]GTP following the method of Geiszt et al. (24) with modifications. Rac or its mutant (6 µM) was loaded with 30 µM [γ-32P]GTP (7 µCi) at 20 °C for 10 min in 8 µL of buffer B (16 mM Tris-HCl, pH 7.5, 0.1 mM dithiothreitol) containing 5 mM EDTA. MgCl2 was then added to a final concentration of 20 mM, and the solution was kept on ice for 1 min. Three microliters of [γ-32P]GTP-loaded Rac or RacQ61L was taken, diluted with buffer B (27 µL) containing 1 mM GTP and BSA (1 mg/mL), and incubated at 20 °C for a given time. An aliquot (5 µL) was taken and filtered onto a nitrocellulose membrane (Schleicher and Schell, BA45), followed by washing three times with 2 mL of cold buffer C (50 mM Tris-HCl, pH 7.7, 5 mM MgCl2). The filters were dried, and the radioactivity was counted in a scintillation counter (Beckman LS6000). For the controls, the experiments were sometimes repeated without the proteins. Far-Western Blotting. Purified p67N (6.6 µg per lane) was electrophoresed in an SDS-polyacrylamide gel (15% gel), and the protein bands were transferred onto nitrocellulose membrane (Schleicher and Schell, BA85). Each band corresponding to p67N (24 kDa) was excised and subjected to far-western blotting. Each piece of membrane was blocked with buffer D (300 mM NaCl and 0.05% Tween 20 in 20 mM Tris-HCl, pH 7.4), and the band was overlaid with Rac or RacQ61L (0.4 µM) at 25 °C for 30 min. For Rac, the protein was preincubated with 100 µM GTPγS at 25 °C for 20 min, and all of the following buffers contained 10 µM GTPγS. After being washed three times (5 min each) with buffer D, each piece of the membrane was incubated with anti-Rac IgG for 12 h at 25 °C. The anti-Rac IgG was a generous gift from Dr. Dave Lambeth (Department of Pathology, Emory University). In a preliminary experiment, we confirmed that there was no difference in anti-Rac IgG
186 Biochemistry, Vol. 42, No. 1, 2003
Miyano et al. Table 1: Effect of Guanine Nucleotides on the Stability of NADPH Oxidase Reconstituted with Wild-Type or Q61L Raca guanine nucleotide GTP GTPγS GDP none
half-life at 37 °C (initial activity)b wild type RacQ61L 12.5 ( 0.8 (2053 ( 20) 11.3 ( 0.7 (1920 ( 69) 9.5 ( 0.2 (1540 ( 20) 3.2 ( 0.1 (1252 ( 13) 6.3 ( 1.4 (513 ( 10)
60.1 ( 5.8 (2798 ( 36) 53.1 ( 2.7 (2571 ( 95)c 60.2 ( 3.7 (2558 ( 62) 57.5 ( 4.1 (2503 ( 42) 68.0 ( 5.7 (2361 ( 29)
Rac or RacQ61L was incubated for 20 min at 25 °C with or without a guanine nucleotide (100 µM) and used in the cell-free reconstitution system containing the same guanine nucleotide at 10 µM. Other conditions were as described in the Experimental Procedures. b The halflives were determined from the data in Figure 1 and other experiments by curve itting to an equation for a first-order reaction. The activities are expressed as nmol of O2- min-1 (nmol of cyt b558)-1. c Preincubation with GTP was performed in the presence of 5 mM EDTA. An aliquot of the mixture was added to the activation mixture containing an excess amount of MgCl2 to achieve a final concentration of 4 mM free Mg2+. a
FIGURE 1: Effect of the RacQ61L mutation of Rac on the thermolability of NADPH oxidase. The reconstitution system contained GTPγS-treated Rac or RacQ61L (6 µM), p67N (6 µM), p47phox (1.3 µM), and purified cyt b558 (0.1 µM). After activation with SDS for 5 min at 25 °C, the mixture was kept at 37 °C for a given time. NADPH was then added to start the enzyme reaction, and O2- generation was measured as described in the Experimental Procedures. Symbols and error bars represent the means ( SD from four determinations. The initial activities with Rac or RacQ61L are shown in Table 1. The graph shows the theoretical curves optimized by curve fitting to a first-order equation.
binding to RacQ61L or wild-type Rac. The membrane was treated with goat anti-rabbit IgG conjugated with peroxidase (Cappel Products) at 25 °C for 2 h and then washed twice with buffer D and once with 20 mM Tris-HCl (pH 7.4) containing 300 mM NaCl. After color development, blots were scanned using a scanner (Epson GT8700) and analyzed with the NIH Image software. RESULTS Effect of the Q61L Mutation on the Stability of NADPH Oxidase. The thermostability of NADPH oxidase activity reconstituted with Rac or RacQ61L was examined. Prior to reconstitution, both Rac and RacQ61L were loaded with GTPγS as a conventional procedure. Following activation of the oxidase, the mixture was incubated at 37 °C for 0-60 min and then assayed for O2- generation. Figure 1 shows the residual activities of the oxidase incubated. The oxidase with RacQ61L was significantly more stable at 37 °C than that with wild-type Rac. After a 30 min incubation, the residual activities of Rac and RacQ61L were 12% and 70% of the initial activities, respectively. As deactivation was found to be a first-order reaction, the half-lives (t1/2) were estimated by curve fitting of the data (Table 1). The halflives of the oxidase with Rac and RacQ61L were 9.5 and 60.2 min, respectively. This indicates a marked ability of RacQ61L to improve the oxidase stability. Effects of Various Guanine Nucleotides on the Oxidase ActiVity. Rac and RacQ61L were pretreated with various guanine nucleotides, and the effects on the stability of the oxidase were examined. For wild-type Rac, the stability of the oxidase varied depending on the guanine nucleotide used
(Table 1). Rac-GTP produced a slightly higher stability (t1/2 ) 12.5 min) than that with GTPγS-treated Rac, and untreated Rac produced a slightly lower stability (t1/2 ) 6.3 min), whereas GDP-treated Rac resulted in very low stability (t1/2 ) 3.2 min). The stability was in the following rank order: GDP , none < GTPγS , GTP. Table 1 also shows the initial activities with different guanine nucleotides. Unexpectedly, the activity with GTPγS-Rac was slightly lower than that with GTP-Rac, and furthermore GDP-Rac produced a substantial activity (61% of the maximal). The activity was in the following order: none , GDP < GTPγS < GTP. In contrast to wild-type Rac, for RacQ61L the stability of the oxidase was not much changed by the nucleotide used. To our surprise, untreated RacQ61L produced a similar, even higher stability (t1/2 ) 68 min) compared with a guanine nucleotide-treated one. Addition of EDTA during the GTP loading did not change the results. On the basis of these results, in the following experiments, Rac was treated with GTP prior to reconstitution, and RacQ61L was used without GTP loading. Stability in the Absence of p47phox. The oxidase can be activated without p47phox when p67phox and Rac are used in excess (15, 16). Therefore, we examined how RacQ61L stabilizes the activity under these conditions. As seen in Figure 2, RacQ61L considerably stabilized the oxidase in the absence of p47phox. The stability with Rac (t1/2 ) 15 min) was improved 8-fold by RacQ61L (t1/2 ) 117 min) (Table 2). The stability was not much increased even when p47phox was added. This indicates that stabilization by RacQ61L occurs independently of p47phox. GTPase ActiVity. To compare the intrinsic GTPase activity of RacQ61L with that of Rac, we performed a filter binding assay using [γ-32P]GTP. The radiolabeled GTP was loaded on Rac or RacQ61L, and GTPase activity was measured at 20 °C. For Rac, GTP was readily hydrolyzed with a rate constant (kcat) of 0.066 min-1. For RacQ61L, only a small population of GTP was exchanged under the conditions, and hydrolysis occurred very slowly with a kcat of 0.003 min-1, which is 20-fold lower than that with Rac. The results indicate that the GTPase activity as well as nucleotide exchange ability is extremely diminished in the RacQ61L mutant. The data, together with Table 1, indicate that
Tightly Bound Phox Triad Keeps NADPH Oxidase on
FIGURE 2: Thermolability of NADPH oxidase reconstituted with a high concentration of p67N plus Rac or RacQ61L. The reconstitution system contained 38 µM Rac (triangles) or RacQ61L (circles), p67N, and purified cyt b558 (0.1 µM). Other experimental conditions were as described for Figure 1. Data represent the means ( SD from four determinations. The initial activities are shown in Table 3. Table 2: Stability of NADPH Oxidase Reconstituted with High Concentrations of p67N and Rac in the Absence and Presence of p47phox a p47phox
Rac
t1/2 at 37 °C (initial activity)b
-p47phox
wild type RacQ61L wild type RacQ61L
14.5 ( 0.9 (1491 ( 46) 116.8 ( 5.8 (2575 ( 43) 24.9 ( 5.3 (2909 ( 78) 128.1 ( 7.3 (3230 ( 62)
+p47phox
a The reconstitution system contained Rac (or RacQ61L) and p67N (38 µM each) in the presence or absence of p47phox (1.3 µM). The halflives were estimated from the data in Figure 2 by curve fitting to an equation for a first-order reaction. b The half-lives (t1/2) are expressed as min and the initial activities are expressed as nmol of O2- min-1 (nmol of cyt b558)-1.
nucleotide exchange is not necessary for RacQ61L to activate the oxidase. Diminished nucleotide exchange ability was previously reported with Rac2 Q61L by Xu et al. (25). Effect of RacQ61L on Kinetic Parameters. The concentration dependence for RacQ61L in cell-free reconstitution was examined and compared with that for Rac. The EC50 values estimated for Rac and RacQ61L were 43 and 16 nM, respectively (Table 3). RacQ61L decreased the EC50 value
Biochemistry, Vol. 42, No. 1, 2003 187
FIGURE 3: Intrinsic GTPase activities of Rac and RacQ61L estimated by a radiolabeled GTP. Rac or RacQ61L (6 µM) was loaded with 30 µM [γ-32P]GTP in the presence of 5 mM EDTA at 20 °C for 20 min. MgCl2 was added to a final concentration of 20 mM to stop the GDP-GTP exchange reaction and start the GTP hydrolysis reaction. The mixture was incubated at 20 °C for a given time in the presence of 2 mM MgCl2, 1 mM GTP, and BSA (1 mg/mL). An aliquot was taken and filtered onto a nitrocellulose membrane. The radioactivity was measured in a scintillation counter. Other experimental conditions were described in the Experimental Procedures.
for p67N 4-fold but did not change the EC50 value for p47phox, and this is consistent with the concept that Rac does not facilitate the interaction of p47phox with cyt b558 (26). The above results show that RacQ61L has a higher affinity for the complex and increases the affinity of p67N in the complex. Table 3 also shows the EC50 values for Rac (Q61L) and p67N in the absence of p47phox. With Rac, the EC50 values for Rac and p67N were 10.4 and 2.9 µM, respectively. In comparison, with RacQ61L the values for RacQ61L and p67N were 2.3 and 1.6 µM, respectively. The results showed that RacQ61L has a higher affinity than wild-type Rac and increases the affinity of p67N in the complex also in the absence of p47phox. However, the affinities are still far lower than those in the presence of p47phox (Table 3). Thus RacQ61L does not substitute for p47phox as an adapter protein. Effect of FlaVin on Thermostability. In the course of the stability experiments, we happened to find that FAD in the incubation mixture influenced the deactivation rate of the
Table 3: Effect of the Q61L Mutation on the EC50 Values for Rac, p67N, and p47phox a Rac
p47phox b
p67N
p47phox
Rac form
EC50 (nM)
Vmax
EC50 (nM)
Vmax
EC50 (nM)
Vmax′
+ + -c -c
wild type Q61L wild type Q61L
43 ( 6 16 ( 6 10400 ( 1900 2300 ( 100
1805 ( 44 2304 ( 124 1255 ( 160 1657 ( 40
69 ( 17 17 ( 4 2900 ( 400 1600 ( 100
2002 ( 83 2342 ( 105 475 ( 31 1245 ( 35
39 ( 16 39 ( 19
1821 ( 164 2379 ( 197
a The cell-free activation was performed as described in the Experimental Procedures. The Vmax values are expressed as nmol of O2- min-1 (nmol of cyt b558)-1. b The activities at [p47phox] ) 0 were not negligible, and therefore the data were fitted to a modified Michaelis-Menten equation, V ) V0 + Vmax/(1 + Km/[S]), where V0 is the activity at [p47phox] ) 0. Furthermore, Vmax′ is defined as follows: Vmax′ ) V0 + Vmax. The V0 values observed for wild-type and Q61L Rac were 323 and 854 nmol of O2- min-1 (nmol of cyt b558)-1, respectively. c The system contained 6 µM p67N with various concentrations of Rac or 6 µM Rac with various concentrations of p67N.
188 Biochemistry, Vol. 42, No. 1, 2003
FIGURE 4: Effect of FAD concentration on the thermolability of NADPH oxidase. After cell-free activation, the mixture was incubated for 30 min at 37 °C in the presence of FAD (0-100 nM). After incubation, the mixture was diluted with an assay mixture containing 10 µM FAD and assayed for superoxide generation. Activities were expressed as the percentages of those without 37 °C incubation. The control activities with Rac and RacQ61L were 2023 ( 20 and 2489 ( 54 nmol min-1 (nmol of cyt b558)-1, respectively. Symbols and error bars represent the means ( SD from four determinations.
oxidase. Therefore, the effect of FAD on the stability of the oxidase complex was investigated in more detail. Following activation, the enzyme was heated at 37 °C for 30 min with various concentrations of FAD and then assayed for O2generation. As shown in Figure 4, the oxidase stability was significantly influenced by the FAD concentration in the incubation. After a 30 min incubation in the absence of FAD, the residual activities with Rac and RacQ61L were 5% and 30%, respectively. When 50 nM FAD was included, they increased 18% and 72%, respectively. The estimated EC50 value with Rac was 28 nM, and that with RacQ61L was 13 nM, suggesting that the affinity of flavin for cyt b558 is increased by RacQ61L. Binding Ability to p67N. As the Q61L mutation increased the affinity of Rac and p67N in the oxidase complex, we next tested whether the mutation increased the interaction between Rac and p67N using far-western blotting. When GTPγS-treated Rac was used, a faint band was observed under the conditions (Figure 5). In comparison, with RacQ61L a more prominent band was observed. The intensity of the RacQ61L band was 3.3-fold higher than that with Rac-GTPγS. Thus, RacQ61L binds to p67N more tightly than the GTP-bound form of Rac. This result is somewhat different from that by Diekman et al. (10), who showed that Rac2Q61L has a slightly lower affinity for p67phox than wild-type Rac by using affinity precipitation. The discrepancy might arise from the difference in the forms of p67phox used. They used full-length p67phox fused with glutathione S-transferase while we used a Cterminal truncated form of p67phox. Effect of Fusion on the Stability. In preceding papers, we found that fusion of p67N with p47N or Rac efficiently
Miyano et al.
FIGURE 5: Detection of interaction between p67N and Rac or RacQ61L by far-western blotting. Purified p67N was electrophoresed in an SDS-polyacrylamide gel, and the protein bands were blotted onto nitrocellulose membrane. The band corresponding to p67N was excised and subjected to far-western blotting as described in the Experimental Procedures. Table 4: Effect of RacQ61L and Fusion Proteins on the Stability of NADPH Oxidasea p67phox form p67N p67N-Rac p67N p67N-RacQ61L p67N-p47N
other proteins p47phox
, Rac p47phox p47phox, RacQ61L p47phox RacQ61Lc
t1/2 at 37°C (initial activity) 12.5 ( 0.8 (2053 ( 20) 35.8 ( 2.6 (2291 ( 44) 63.6 ( 16.0 (2952 ( 71)b 86.3 ( 21.2 (3049 ( 12) 184.0 ( 33.0 (3501 ( 30)
a The reconstitution system contained p67N and RacQ61L or the p67N-RacQ61L fusion protein (6 µM each), p47phox (1.3 µM), and purified cyt b558. Other conditions were as described in the Experimental Procedures. Half-lives (t1/2) are expressed as min and initial activities are expressed as nmol of O2- min-1 (nmol of cyt b558)-1. b When p47N was used, the half-life was 84.0 ( 8.8 min and the activity was 3085 ( 4 nmol of O2- min-1 (nmol of cyt b558)-1. c p67N-p47N (2 µM) and RacQ61L (6.8 µM) were used in the reconstitution.
stabilizes the oxidase. Therefore, here we applied a fusion technique to the system with the RacQ61L mutant. When fused p67N and RacQ61L were used in the system, the stability of the oxidase was slightly higher (t1/2 ) 86 min) than that with the individual components (t1/2 ) 64 min) (Table 4). However, the effect of fusion was not as dramatic as that observed with p67N and Rac (18). This suggests that RacQ61L and the fusion protein share a common mechanism for stabilization, namely, enhancement of the interaction between p67N and Rac. In contrast, when RacQ61L was used with the p67Np47N fusion protein, the oxidase was extremely stabilized. The stability (t1/2 ) 184 min at 37 °C) was the highest level of all the approaches we tested, including fusion and crosslinking (17, 18, 23). The effectiveness for stabilization was in the following rank order: p67N, Rac < p67N-Rac < p67N, RacQ61L < p67N-RacQ61L , p67N-p47N, RacQ61L. Of interest is the fact that the stability with
Tightly Bound Phox Triad Keeps NADPH Oxidase on RacQ61L was higher than that with p67N-Rac, a covalently bound form of the protein pair. The remarkable stabilization with p67N-p47N and RacQ61L should be due to the formation of a very stable ternary complex formed among p67N, p47N, and RacQ61L. DISCUSSION The Q61L mutation in small GTPases was originally discovered in a Ras variant from human lung carcinoma cells (27) and assumed to be the cause of its oncogenic nature. RasQ61L has an impaired GTPase activity as with RasG12V, which might lead to continuous activation of Ras and continuous binding to Raf and, consequently, to nonregulated cell proliferation through intracellular signal transduction. The Q61L mutant of Rac has often been used as a constitutively active form of Rac, i.e., as a persistently GTPbound form. In this study, we used RacQ61L in a pure reconstitution system of NADPH oxidase and found that it remarkably improved the stability of the oxidase. The data showed the following properties of RacQ61L and the RacQ61L-containing system: (i) the affinity of Rac and p67N in the complex is increased by the mutation, (ii) the stabilization occurs in either the presence or absence of p47phox., (iii) RacQ61L has a diminished guanine nucleotide exchange ability, and (iv) in combination with p67N-p47N, RacQ61L produced an extremely stable activity of the oxidase. As expected, RacQ61L had an impaired GTPase activity, but this nature does not seem to be a major reason for the improved stability, because GTPγS-loaded Rac did not stabilize the oxidase as effectively as RacQ61L. Rather, a diminished nucleotide exchange, i.e., a suppressed GTP release (25), might contribute to the stability. Another possibility is a conformational change in the Rac structure by the RacQ61L mutation. The crystal structure of wild-type Rac revealed that Gln at position 61 is located close to switch I (28), which interacts with p67N (13). Leu substituted for Gln-61 may have contact with switch I (particularly Pro-34) and move the loop or induce a conformational change in switch I, which is favorable to p67N binding. Xu et al. reported that Rac2Q61L loses the ability to bind GDS (25) and reverses the inhibitory effect of mutations in the effector domain such as D38A (29). These facts suggest that a drastic conformational change occurs in RacQ61L and that RacQ61L has a partially different conformation from that of RacG12V, which shares the properties of low GTPase activity and low responsiveness to GAP (29). Thus the stabilization of the oxidase activity by RacQ61L may be due to a conformational change in RacQ61L induced by the mutation, which results in an enhanced binding between RacQ61L and p67N and leads to the complex stabilization. In the present study, we have established a remarkably stabilized NADPH oxidase by combination of RacQ61L and a fusion protein p67N-p47N. The stability is at the highest level of all of the approaches we tested (17, 18, 23). Our recent studies showed that the p67N-Rac or p67N-p47N fusion protein stabilizes the oxidase and suggested that the orientation of the activation domain of p67phox at the correct position is centrally important for activation (23). The data predicted that if the N-termini of p47N and Rac were both
Biochemistry, Vol. 42, No. 1, 2003 189 linked to the C-terminus of p67N, stabilization would be more synergistically improved. However, the construction of a triple fusion protein of such an arrangement is impossible using genetic engineering. Fortunately, in the present study we found a tight binding between RacQ61L and p67N. This provided us with the opportunity to overcome the experimental limitations and produce a mostly stable ternary complex among p47N, p67N, and Rac. It is widely accepted that interaction between cytosolic phox proteins and cyt b558 is a major process in the oxidase activation. In addition, the stoichiometry of the protein components in the complex formation has been postulated (5, 6, 30). However, Cross et al. reported that the activation process is primarily catalytic and not through the formation of a stoichiometric complex (31). Furthermore, Quinn et al. showed that complex formation is required but not enough for activation (32). Thus, the contribution of the stoichiometric complex to the duration of the activation has not been established. Our present results show that complex formation is centrally important in the duration of the oxidase activity. One of the most interesting findings in this paper is that the oxidase stability is greatly influenced by FAD. FAD increased the stability 4-fold at a nanomolar order of concentration. The finding suggests that FAD, a chromophore of NADPH oxidase, has a role in maintaining the oxidase complex. The presence of FAD on gp91phox contributes to the conformational stability of the molecule. We also found that the flavin concentration required for stabilization is lowered by the RacQ61L mutant. It is tempting to speculate that p67N, which is tightly bound to gp91phox through RacQ61L, may prevent FAD dissociation either directly or indirectly. The present experiments showed that GDP-bound Rac can produce substantial activity although the stability is low. Canonical small GTPases bind to their effector proteins as a GTP-bound form (33) although there are atypical small GTPases that bind to their effector proteins in GDP form (e.g., ran and rab) (34, 35). With regard to the phagocyte NADPH oxidase, the Rac form that activates the oxidase is controversial. Many groups have reported that the GTP form is necessary for the activation (36-39), but some groups have demonstrated that the GDP form is also capable of activating the oxidase (40-42). In addition, it has recently been demonstrated that the interaction between p67phox and cyt b558 is enhanced by either the GTP or GDP form of Rac (43). Further studies will be required to clarify this point. In summary, using a mutant of Rac and a fusion protein, we have accomplished the production of an extremely stabilized NADPH oxidase in a pure reconstitution system. This indicates that a tightly bound ternary complex among p67phox, Rac, and p47phox is very effective in maintaining the oxidase activity and confirms the concept that the longevity of the activated enzyme requires continuous association of these cytosolic components as suggested in our earlier study (17). Finally, we emphasize that this simple and efficient method to stabilize the oxidase may provide a useful tool for elucidating the nature of the activated complex, which is transient in both cells and cell-free systems.
190 Biochemistry, Vol. 42, No. 1, 2003 ACKNOWLEDGMENT The authors are indebted to Miharu Mori, Teruaki Nagasawa, Shinya Ogasawara, Mamoru Tanaka, Kazuaki Komeima, and Kyoji Watanabe (Department of Applied Chemistry, Ehime University) for technical assistance. We are grateful to Dr. Ryo Morishita and Ozawa Akihiko (Department of Applied Chemistry, Ehime University) for instructions for the RI experiments. We also thank Drs. Yaeta Endo and Koichi Kato (Department of Applied Chemistry, Ehime University) for the use of instruments and Drs. Chang-Hoon Han and Dave Lambeth (Department of Pathology, Emory University) for valuable discussions. REFERENCES 1. Babior, B. M. (1978) N. Engl. J. Med. 298, 659-668. 2. Rossi, F. (1986) Biochim. Biophys. Acta 853, 65-89. 3. Thrasher, A. J., Keep, N. H., Wientjes, F., and Segal, A. W. (1994) Biochim. Biophys. Acta 1227, 1-24. 4. Lambeth, J. D., Cheng, G., Arnold, R. S., and Edens, W. A. (2000) Trends Biochem. Sci. 25, 459-461. 5. DeLeo, F. R., and Quinn, M. T. (1996) J. Leukocyte Biol. 60, 677-691. 6. Babior, B. M. (1999) Blood 93, 1464-1476. 7. Sumimoto, H., Kage, Y., Nunoi, H., Sasaki, H., Nose, T., Fukumaki, Y., Ohno, M., Minakami, S., and Takeshige, K. (1994) Proc. Natl. Acad. Sci. U.S.A. 91, 5345-5349. 8. Leto, T. L., Adams, A. G., and de Mendez, I. (1994) Proc. Natl. Acad. Sci. U.S.A. 91, 10650-10654. 9. McPhail, L. C. (1994) J. Exp. Med. 180, 2011-2015. 10. Diekmann, D., Abo, A., Johnston, C., Segal, A. W., and Hall, A. (1994) Science 265, 531-533. 11. Ahmed, S., Prigmore, E., Govind, S., Veryard, C., Kozma, R., Wientjes, F. B., Segal, A. W., and Lim, L. (1998) J. Biol. Chem. 273, 15693-15701. 12. Koga, H., Terasawa, H., Nunoi, H., Takeshige, K., Inagaki, F., and Sumimoto, H. (1999) J. Biol. Chem. 274, 25051-25060. 13. Lapouge, K., Smith, S. J. M., Walker, P. A., Gamblin, S. J., Smerdon, S. J., and Rittinger, K. (2000) Mol. Cell 6, 899-907. 14. Nisimoto, Y., Freeman, J. L. R., Motalebi, S. A., Hirshberg, M., and Lambeth, J. D. (1997) J. Biol. Chem. 272, 18834-18841. 15. Freeman, J. L. and Lambeth, J. D. (1996) J. Biol. Chem. 271, 22578-22582. 16. Koshkin, V., Lotan, O., and Pick, E. (1996) J. Biol. Chem. 271, 30326-30329. 17. Tamura, M., Takeshita, M., Curnutte, J. T., Uhlinger, D. J., and Lambeth, J. D. (1992) J. Biol. Chem. 267, 7529-7538.
Miyano et al. 18. Miyano, K., Ogasawara, S., Han, C.-H., Fukuda, H., and Tamura, M. (2001) Biochemistry 40, 14089-14097. 19. Cross, A. R., and Curnutte, J. T. (1995) J. Biol. Chem. 270, 65436548. 20. Han, C.-H., Freeman, J. L. R., Lee, T., Motalebi, S. A., and Lambeth, J. D. (1998) J. Biol. Chem. 273, 16663-16668. 21. Nisimoto, Y., Motalebi, S., Han, C.-H., and Lambeth, J. D. (1999) J. Biol. Chem. 274, 22999-23005. 22. Tamura, M., Tamura, T., Burnham, D. N., Uhlinger, D. J., and Lambeth, J. D. (1989) Arch. Biochem. Biophys. 275, 23-32. 23. Ebisu, K., Nagasawa, T., Watanabe, K., Kakinuma, K., Miyano, K., and Tamura, M. (2001) J. Biol. Chem. 276, 24498-24505. 24. Geiszt, M., Dagher, M.-C., Molnar, G., Havasi, A., Faure, J., Paclet, M.-H., Morel, F., and Ligeti, E. (2001) Biochem. J. 355, 851-858. 25. Xu, X., Wang, Y., Barry, D. C., Chanock, S. J., and Bokoch, G. M. (1997) Biochemistry 36, 626-632. 26. Kleinberg, M. E., Malech, H. L., Mital, D. A., and Leto, T. L. (1994) Biochemistry 33, 2490-2495. 27. Yuasa, Y., Srivastava, S. K., Dunn, C. Y., Phim, J. S., Reddy, E. P., and Aaronson, S. A. (1983) Nature 303, 775-779. 28. Hirshberg, M., Stockley, R. W., Dodson, G., and Webb, M. R. (1997) Nat. Struct. Biol. 4, 147-152. 29. Xu, X., Barry, D. C., Settleman, J., Schwartz, M. A., and Bokoch, G. M. (1994) J. Biol. Chem. 269, 23569-23574. 30. Uhlinger, D. J., Tyagi, S. R., Inge, K. L., and Lambeth, J. D. (1993) J. Biol. Chem. 268, 8624-8631. 31. Cross, A. R., Erickson, R. W., and Curnutte, J. T. (1999) Biochem. J. 341, 251-255. 32. Quinn, M. T., Evans, T., Loetterle, L. R., Jesaitis, A. J., and Bokoch, G. M. (1993) J. Biol. Chem. 268, 20983-20987. 33. Cotteret, S. and Chernoff, J. (2002) Genome Biol. 3, 1-8. 34. Pfeffer, S. R. (1992) Trends Cell Biol. 2, 41-46. 35. Stewart, M., Kent, H. M., and McCoy, A. J. (1998) J. Mol. Biol. 277, 635-646. 36. Heyworth, P. G., Knaus, U. G., Settleman, J., Curnutte, J. T., and Bokoch, G. M. (1993) Mol. Biol. Cell 4, 1217-1223. 37. Abo, A., Webb, M. R., Grogan, A., and Segal, A. W. (1994) Biochem. J. 298, 585-591. 38. Fuchs, A., Dagher, M.-C., Jouan, A., and Vignais, P. V. (1994) Eur. J. Biochem. 226, 587-595. 39. Kwong, C. H., Malech, H. L., Rotrosen, D., and Leto, T. L. (1993) Biochemistry 32, 5711-5717. 40. Toporik, A., Gorzalczany, Y., Hirshberg, M., Pick, E., and Lotan, O. (1998) Biochemistry 37, 7147-7156. 41. Bromberg, Y., Shani, E., Joseph, G., Gorzalczany, Y., Sperling, O., and Pick, E. (1994) J. Biol. Chem. 269, 7055-7058. 42. Grizot, S., Faure, J., Fieschi, F., Vignais, P. V., Dagher, M. C., and Pebay-Peyroula, E. (2001) Biochemistry 40, 10007-10013. 43. Dang, P. M.-C., Cross, A. R., and Babior, B. M. (2001) Proc. Natl. Acad. Sci. U.S.A. 98, 3001-3005. BI0269052
